At the first meeting of SAFE (9 June 2009) I volunteered to contact relevant agencies to help clarify some of the points made during the public meeting. This was in the context of a report stating that 300 homes in Sidmouth were at risk of being flooded, that areas of Sidmouth were below sea level and could be inundated, with loss of trade to shops, etc.

What follows is a brief and somewhat unstructured summary - I have taken 'at face value' what I have been told and tried to consolidate it into some sort of coherent whole. If some of the data I was given is wrong, then my interpretations may be in error.

Crucially, given the concerns raised on 9 June, it is to be noted that joint probability modelling for combined tidal/fluvial flooding events is not undertaken as a standard part of EA work in producing the National Flood Zone maps. Also, I can make no comment as yet on any of the reports held by EDDC - they only agreed to pursue my requests for information after some weeks of email exchanges and even as of January 2010 have not provided some of the information I requested.

There seem to be several crucial issues for the future:

Ascertaining the true rate of recession of the cliffs. I am amazed that no proper monitoring has been undertaken over the last 20 or 30 years. Even a few dozen pegs in the ground and a tape measure would have provided valuable information. I suggested a study of old (dated) aerial photos if any exist, or satellite data held by government? This seems worth pursuing - something that EDDC or EA could do? The old cliff path would presumably provide a datum reference line. Presumably there are accurate maps of where the old/new coastal paths were/are located?

Ascertaining why the sea defences to the west caused the beach to the east to disappear. This really needs to be understood in as much detail as possible. It might point the way to a simpler solution to protecting the cliffs to the east?

Is there any possibility of a large cliff fall into the Sid upstream of the Alma bridge? This could cause inundation of parts of the town centre if it occurred during a period of severe flooding on the Sid.

Joint probability modelling is not undertaken routinely by the EA but might be relevant if it is supposed that the risk to the town results primarily from a severe south-east gale and tidal flow together with severe flooding on the Sid. It is unclear whether the combined probability of tidal/river events pose a serious risk to Sidmouth - over and above each risk taken alone. If not, then recession of the cliffs remains a matter of principal interest only in respect of loss of cliff top properties (eventually) and maybe loss of the Alma Bridge. Economic analysis might need to be undertaken to help justify the costs of any proposed major works.

It may be helpful to obtain the EA flood zone maps and overlay them onto street maps to see which areas are thought (from imperfect modelling) to be at risk in the town centre. This is something the EA could produce, maybe at no cost. Some information is on their website and indeed on the EDDC website showing areas at risk down to street level.

I spoke to the EA in Exeter on 10 June enquiring about a supposed recent EA report stating that 300 houses are at risk of flooding in Sidmouth. Another EA employee called me back on 17 June and again on 27 June to provide further information after I had explained the context of my enquiries on behalf of SAFE.

There appears to be no EA report specifically related to Sidmouth. The EA undertakes its National Flood Zone Mapping exercise to identify areas at risk from fluvial flooding (from rivers overtopping their banks, so called 'out of bank flows') and tidal flooding (from the sea). These studies do not map areas at risk from surface water problems.

Several of the areas flooded in the UK in recent years have suffered because of high local rainfall onto land served by inadequately sized surface water drains. This is different from 'out of bank' flows from rivers and can occur miles from the nearest river or floodplain.

The purpose of the EA National Flood Zone project is as a broad brush 'first guess' as to where flooding is most likely, based on land topography. The primary use is for land use planning - for example, suggesting where a new town or housing estate might best be situated to minimise risks. The methodology does not include any consideration of flood defences that may have been constructed (unless these are large enough to show on the land topography assessed by laser measurements from aircraft, which is how the system works) and does not include any consideration of drainage capacity. Therefore it is indicative of flood risks but cannot provide accurate data for a small area. There have been some examples of where houses have been 'blighted' for insurance purposes because they were in the same postcode as houses known to suffer flooding - again the methodology cannot usually help with these issues. I am putting some specific examples onto my website.

For EA purposes 'Sidmouth' is assessed as the extent of the urban community, it is not defined by any historical parish boundaries or similar. Taking this as the definition of 'Sidmouth' the EA have provided figures for houses potentially at risk in their Zone 2 which is an extreme flood outline - defined as once in 1000 years. For tidal flooding, there are 263 homes at risk (which may be where the quoted figure of 300 comes from) and for fluvial flooding 627 houses (these may be 'properties' rather than all houses).

Within Zone 3 (defined as 1 event in 200 years for tidal flooding and 1 event in 100 years for fluvial flooding) there are again 263 properties at risk (tidal) and 498 for fluvial flooding. Note that the tidal figure has not changed from the 1000-year event. This is not unexpected: the most severe tidal event in 200 years may well be the most severe expected in 1000 or even 2000 years. Short of a local tsunami, the sea can only get so rough. Severe rainfall has a broader expected range - whilst 100mm in an hour onto already saturated ground might be the requirement for a 100 year flood there is little to prevent 200mm in an hour - giving a different 1000-year event, for example.

Indeed, Sidmouth (and especially Sidford) might have experienced this if the freak storm that centred on the Otter valley and affected Ottery St Mary in late 2008 was instead centred upon the Sid valley. The storm simply stayed in one small area for hours - and if that area had been anywhere in the Sid valley the river would have responded accordingly - and far faster than the Otter did owing to the steeper gradients.

These figures of hundreds of properties at risk seem alarmingly high but may take little or no account of the deep engineered channel in which the Sid is constrained to flow for the last few hundred metres before it reaches the sea. This should (apparently) have been picked up by the laser measurements and incorporated into the modelling - but it seems all rather uncertain (if not imprecise) and should only be used as a rough indicator of areas and (thereby) numbers of properties at risk.

For Sidford, there are no houses at risk from tidal flooding - the tidal range of the Sid is quite short, reflecting the steep gradient of the valley, and it does not come anywhere near Sidford. For fluvial flooding, there are 12 properties at risk (1 event per 1000 years) and only 9 properties at risk (1 event per 100 years). I find these figures remarkably low - given that the Sid washed to the end of Packhorse Close in 1997 - and it is not the only road that was underwater. The modelling may not accurately take into account the Packhorse Bridge that acts as a 'bottleneck' and is primarily responsible for how the river behaves either side of the A3052.

Recent changes to the land at Hamilton Garage (for example) may have substantially increased the risk to properties in Hamilton Close but reduced the risk to those in Packhorse Close slightly. Such changes can only be understood in the context of local alterations to land levels, local flood defences, etc. None of these can be incorporated into the EA maps of 'areas potentially at risk' because neither the methodology of measurement of land levels nor the modelling of river flows that utilise the land level data can incorporate such fine detail.

Therefore, on my understanding, the EA have some data on areas of the town 'at risk' and these can be overlaid on road maps to assess the number of properties at risk. This may be how the EDDC maps are created. However, whether the river would behave exactly as predicted by the (imperfect) computer models is a moot point - and the areas actually flooded could be substantially determined by local topography, local flood defences, etc. A serious cliff fall into the Sid that served substantially to block its flow at a crucial point - just upstream of Alma Bridge for example, could much alter the behaviour of the river and its effect on town centre properties. This would be compounded if central Sidmouth has inadequate surface water drainage.

Based on events in 1997, my own guess is that Sidmouth is less at risk than the EA figures might suggest (at least for fluvial flooding and assuming no substantial cliff fall into the Sid) and areas of Sidford may be significantly more at risk - but it is only a guess. I would need a far better understanding of how the data was collected and processed to make any further assessment.

The final important point is that none of the EA calculations include joint probability modelling - that is, the flooding that might be expected if a severe tidal event coincided with a severe flood on the Sid. This modelling could in principle be undertaken (at cost?) but would be subject to the same if not perhaps greater uncertainties. Relevant parameters would include the probability of the most serious storm from the sea (a south-easterly gale and high tide?) combined with severe rainfall. The joint probability might be less than for a south-westerly gale - because south-west winds generally bring more rain than do those from the south-east.

Southwest water (SWW) has also been helpful. An employee phoned me back twice and again provided relevant data and explanations.

Internal flooding of houses occurs when water in sewers can find an escape route via ground-floor toilets (etc) in preference to being constrained within the sewer system. It is a problem in towns and cities where there are old so-called 'combined' sewers. These carry both rainwater (surface water) and foul water from houses and other premises. The problem is simply one of hydraulic design of the sewers - at certain points they are inadequately sized and under storm conditions internal flooding is a predictable event. The process is called 'hydraulic overload' but affects (apparently) only one property in Sidmouth - and this is well away from the seafront. In all of Devon, Cornwall and Somerset there are only 100 affected properties and these are listed on the DD5 register (whatever that is) and searches are undertaken as part of the HIPS package for house sellers (as q21 on form CON29DW). The criteria for inclusion include that an event is expected once in every 20 years - this is set by OFWAT. Improvements to sewers are expected to reduce the number of houses potentially affected - but it can be far cheaper to protect a few houses than reengineer miles of sewers.

In some areas (including Woolbrook in Sidmouth) overload can cause metal covers in the road to lift - this is an accepted 'safety valve' mechanism and short of laying a new and larger sewerage and/or surface water system, little can be done about it. For individual houses that are potentially affected, non-return valves can be fitted to prevent water surging back into the house. But it is not an issue in Sidmouth.

Internal flooding can also occur in properties that are not normally 'at risk' if there are blockages in sewers - again improvements to and relining of sewers is reducing the incidence across the UK.

Contrary to views expressed at the meeting on 9 June, no houses in Sidmouth are below sea level. They might appear to be so when viewed from the Esplanade, but a better vantage point is the river wall by the swimming pool. It is a greater vertical distance down to the Sid than down into town - and the Sid flows reliably into the sea.

Also, before the days of the new sewerage system, all properties and all streets in Sidmouth drained quite happily into the combined sewers from which everything flowed out to sea (via the macerator on the Ham that chewed up the solids a bit).

There is therefore no chance (yet) of the sea crashing over the river wall, inundating the town and somehow refusing to leave. This was suggested as a possibility on 9 June. If sea levels rise (in 50 or maybe 200 years) Sidmouth may well have more serious problems to address, and the town could not easily be protected - defences might include a higher Esplanade and river wall (if only to prevent storms overtopping the defences). Some of the town centre area (if it became actually below sea level owing to sea level rise) might one day be subject to prolonged flooding until such time as it could be pumped out again - the sewerage system being incapable of dealing with storm surges even today. At the present time (so I believe) surges are dealt with by releasing some water directly to the sea (this would be untreated but heavily diluted sewage). Such 'gravity aided' release would no longer be possible if sea level had risen appreciably - water does not flow uphill.

In the even longer term, maybe a new Esplanade at Sidford?

I have emailed EDDC to ask about the availability (under the FoI Act) of all the reports produced on sea defences for Sidmouth. In particular I asked for details of the correspondence that (apparently) predicted problems with the beach to the east of the town were the rock islands (etc) to be constructed to the west.

These are important issues - if it was predicted that the beach to the east would disappear (leaving the cliff base more exposed) then data may exist showing why this was to happen and even suggesting how it the beach might be caused to return. Completing the 'as envisaged' construction might have had a different effect on the beach than the incomplete design that was (again, apparently) all that was constructed.

EDDC have been much slower to respond helpfully than either EA or SWW - but I have asked more of them. As of January 2010 I am still telephoning Keith Steel (Engineers Dept) at regular intervals. He keeps saying he is trying to get EDDC's Finance Dept to respond.

Stephen J Wozniak.
Sidmouth.

back to top of section